1.01

Financial Impact of Lexington’s Voluntary Desegregation (METCO) Program October 2016 Mark Andersen, Ph.d., CFA

This paper is one of a series on issues posed by Lexington’s participation in voluntary

desegregation (METCO) under the Massachusetts Racial Imbalance Act.

Lexington’s 300th Anniversary Celebration reflects heightened diversity in the 21st Century 1

Executive Overview

Lexington Public Schools participate in a voluntary desegregation program (METCO ) 2

under the Massachusetts Racial Imbalance Act, which permits 250 Boston students to

attend Lexington schools at no cost to Boston taxpayers and with limited subsidy from

the Commonwealth of Massachusetts. This report estimates that the METCO program

costs Lexington taxpayers $5.6 million per year, consisting of $4.2 million in operating

costs and $1.4 million in capital expenses. These figures are conservative, excluding

factors such as increased special education use by METCO students which impact both

operating costs and space needs. As Lexington taxpayers confront surging enrollment

1 Photograph from Lexington 300 Celebration FB page: goo.gl/Rk2ioF 2 https://en.wikipedia.org/wiki/METCO

which drives budgetary increases and capital expansion at all school levels, and even a

re-evaluation of neighborhood schooling in Lexington, taxpayers should understand the

financial consequences of participation in METCO.

METCO was initiated in 1966 to help Boston’s African-American students at a time when

few minority students attended Lexington schools. In the beginning, the METCO

program was financially secure, with Lexington charging full tuition for METCO students.

As state financial support eroded, METCO advocates offered an “empty seats” cost model

suggesting that educating METCO students cost communities little. This report mirrors

prior studies in Brookline, Newton and Lincoln which demonstrated that METCO is

expensive for receiving districts, while the sending district (Boston) contributes nothing.

Because Lexington has crowded schools and numerous classrooms at maximum sizes

permitted by collective bargaining agreements, the effective cost of METCO is perhaps

larger than ever before. But in 2016, the state only provides $1.6 million “grant” funds for

transporting and educating 254 students including their special needs (a cost which was

hardly a major driver in 1966).

Lexington’s METCO program is the third largest in the state, with Lexington receiving 8%

of Boston’s total METCO students even though Lexington represents less than 1% of the

metropolitan area population. Most Boston suburbs do not participate in METCO. Some

argue that Lexington is an affluent town, and should share its largesse with less

privileged students from Boston. However, METCO students represent a variety of ethnic

groups (African-American, Hispanic, Asian) and are not chosen on the basis of economic

need. In addition, Lexington’s excellent schools have attracted so many students that 3

local taxpayers contribute disproportionately to education according to metrics such as

households per public school student. And lastly, today Boston taxpayers pay a lower

residential tax rate than Lexington, and contribute nothing to the regular and special

education costs of Boston METCO students. 4

3 One report suggests that METCO students are financially more secure than the average Boston students. METCO students are 50% low income compared with Boston students (all races) at 74% low income. METCO Inc. uses a waitlist system for referrals, which likely gives advantage to well-connected over transient families. The referral and acceptance process also poses logistical hurdles which may discourage transitional family participation. METCO Merits More , http://prrac.org/pdf/METCOMeritsMore.pdf, p.9,11. 4 Nor are Lexington students eligible to attend Boston’s magnet schools. Lexington’s residential tax rate is $14.60 per thousand , Boston’s tax rate is $11.00 per thousand (https://goo.gl/Wj3OPY) Not only

2

What alternative uses exist for $5.6 million per year? It could be used to reduce the

average household tax bill by $549 per year, or build several proposed projects (a police 5

station, a fire station, and a visitor center), or pay for half of a new high school with

updated science and engineering labs, or institute the proposed $2.3 million elementary

foreign language while supporting accelerated math programs.

While Lexington struggles to accommodate its increasing enrollment, taxpayers should

evaluate whether the original conditions warranting METCO (such as lack of diversity in

Lexington or Boston schooling choices) have passed. Absent increased funding for 6

METCO, Lexington would need to close this $5.6 million per year financial gap and

address capacity on its own. Possible options seem to include phasing out METCO,

downsizing the program, or eliminating METCO-specific staff positions. However,

substantial operating and capital relief can only be obtained by phasing out METCO or

substantially reducing program size.

Introduction

This paper estimates the current fiscal impact of educating Boston students in 7

Lexington, using an approach which parallels approaches used in Brookline, Newton and

Lincoln to demonstrate that these few districts are now providing multi-million dollar

program subsidies unbeknownst to most taxpayers. METCO Inc. directors provide a

narrative that students occupy “empty seats”, with little cost to Lexington. This report 8

does Boston have a lower rate, but Boston residents enjoy a residential tax exemption reducing each resident’s tax bill by $1,962 (https://goo.gl/Q8Zqxb). 5 Lexington has 10,195 households in 2016 per Town of Lexington, Financial Summit I , 10/6/2016, p23. 6 METCO was planned to be a short term program lasting five years at most. A Qualitative Examination of Desegregated Education Through the METCO Program , Kahris Dianne White-McLaughlin, 1996, p114. Also Ruth Batson in “Boston’s METCO: What To Do Until the Solution Arrives”, City , 1971. For a discussion of Lexington’s rapidly increasing diversity, see How Lexington’s Voluntary Desegregation (METCO) Program Violates the Massachusetts Racial Imbalance Act (RIA) , by Mark Andersen. 7 The term “Boston students” and “METCO students” are used interchangeably in this paper to refer to students who live in Boston and participate in METCO, and without this voluntary exchange program would be the responsibility of the Boston Public Schools. 8 “Empty seats” is repeated consistently and conceptually dates from 1968 or earlier (see “Metropolitan Cooperation in Education”, by Efrem Sigel and Gary Jonas, Journal of Negro Education , Vol 39, No 2, Spring 1970, p151, 154). For examples see Wayland HS (http://whsmetco.blogspot.com/) visited 10/15/2016; METCO’s Enroll Today tab accessed 8/2/2016 (since changed) states: “districts request students to fill empty spaces in their schools”; METCO in Brookline (https://goo.gl/ymqe0S, https://prezi.com/-ywjthlw8fxt/metco/), in Newton (http://www.newton.k12.ma.us/Page/866) in

3

provides clear evidence contradicting this political narrative, specifically in Lexington, but

also applicable in most receiving districts. In 2016, Lexington does not have empty seats,

but instead is scrambling to obtain land and install modular classrooms to accommodate

the current and growing student population, all the while underwriting voluntary

desegregation with Boston. 9

Qualitative Supporting Evidence

Statements from METCO proponents suggest host communities face significant expense,

contradicting an “empty seats” cost model. For example, the Massachusetts DoE website

underscores the lack of full funding for the METCO program: 10

[G]iven the low reimbursement and the present level funding of the program, it is

unclear exactly how a school district could join without additional overall funding to

the program itself. 11

The Pioneer Institute’s white paper “Metco Merits More” states that the METCO program

is subsidized by receiving districts (my italics):

“[S]tate dollars provided for METCO fall far short of actual per-pupil costs. This

means that suburban METCO communities, increasingly struggling with budget

concerns themselves, are underwriting a share of the costs related to the

program”. 12

If experts agree that METCO costs receiving districts significant money, we must then ask

ourselves “How Much?”

Foxboro (https://goo.gl/xPwd7L), Lily Geismer, Don’t Blame Us , Princeton University Press, 2015, p218. While “empty seats” is often found on these websites and promotional literature, Lexington’s town reports from 1966-1970 demonstrate that METCO students were introduced while Lexington was undertaking capital projects to relieve crowding. 9 How Lexington’s Voluntary Desegregation (METCO) Program Violates the Massachusetts Racial Imbalance Act (RIA) , Mark Andersen, October 2016 discusses the Racial Imbalance Act (RIA), which legally enables the METCO program. RIA is a desegregation law, and permits the transfer of students from predominantly non-white districts to racially isolated (predominantly white) districts and schools. This paper argues that Lexington is no longer eligible to be a receiving district because Lexington now is racially imbalanced with more than 50% non-white students. 10 https://goo.gl/p30Dlq 11 http://www.doe.mass.edu/metco/faq.html?section=b, Accessed 8/8/2016. 12 Metco Merits More , pp3-4.

4

Financial Analysis

When METCO was launched, participating districts were told there would be no financial

cost. During fifty years of program administration, the Massachusetts Department of 13

Education has distinguished “Direct Program Costs” (METCO specific line items) from

“School Personnel Costs” (offsetting school charges such as teaching staff) to calculate

total “Incremental Instructional Costs” for METCO. With each district independently

setting tuition rates and establishing and expanding direct services for METCO students,

resulting reimbursements varied widely up through at least 1990. In recent years, the 14

state has standardized reimbursement per pupil based on formula allocation, and

receiving districts have lost the ability to set tuition based on local expenditures. Today,

receiving districts accept reimbursement rates as published, and simply fill out a grant

application to explain how they will spend METCO grant funds. Using grant and budget

data, this report shows that funds provided by the state do not fully cover Lexington’s

transportation and METCO specific services, and typically receiving districts must also

fund the entire cost of common educational services for METCO students.

Budgetary Analysis of Operating Expenditures

METCO Grant Funding Process

Since 2006, Lexington has received formula funding for METCO transportation and

services totalling between $1.3 and $1.6 million per year for about 250 students. For 15

2016, every district receiving METCO students (“receiving districts”) can apply for a grant

with a preset amount based on two formulaic numbers: one for education and one for

transportation. For Lexington, these formulas are based on METCO enrollment as of

10/1/2015 (254 students). The educational reimbursement for FY17 has been proposed 16

13 https://goo.gl/8oVR79. Also documented in Lexington school committee minutes and contemporary letters to the newspaper editor. In 1968, Lexington voluntarily cut its METCO tuition rate in half, but still charged tuition it set at average cost per pupil, on top of receiving transportation support from external funds. 14 METCO Archives, Box 7, Folder 15, Minutes, 1991 documents METCO Inc. objections to varied reimbursement. These and other folders from the METCO archives are held in the Archives & Special Collections Department at Northeastern University Libraries. 15 http://www.doe.mass.edu/metco/funding.html 16 http://www.doe.mass.edu/grants/2017/317/Allotment.pdf

5

by Massachusetts to be $3,925 per pupil, totalling $1,032,510. Transportation 17

reimbursement also reflects a state calculation, and Lexington’s participation earns it

$525,094 in transportation funds.

The historical record of these grant applications speaks to the funding of this program.

In prior years, Lexington grant funds were used both to fund METCO specific services

(direct services) and a also to offset indirect educational costs (teachers and materials).

After years of flat funding, and a desire to provide directed METCO resources, FY2017’s

proposed $3,925 per pupil is so limited that all funds are typically for METCO direct

services, leaving zero offset for common educational expenses. Data from the 2016

METCO Program Report shows that Lexington’s use of funds is similar to other districts: in

almost no districts are funds now available to offset indirect costs (shared teachers and

materials), as shown in Figure 1 rightmost column.

17 http://www.doe.mass.edu/grants/2017/317/. “Proposed” because the Governor can use 9C authority to reduce the METCO budget later, as has been done twice in recent years. http://massbudget.org/report_window.php?loc=9CAuthority.html.

6

Figure 1: Expenditure of METCO Funds by School District18

Lexington Use of METCO Grant Funds

The Lexington Superintendent’s Recommended Budget (1/5/2016) summarizes

Lexington’s planned use of grant funds. This budget was prepared based on an earlier

estimate of $1.47 million for FY2017, and divides between transportation spending of

$590,763 and direct services of $897,800. This first recommended budget proposed to

spend slightly more on METCO specific services than the initial METCO state grant, with

zero funds used to offset common educational expenses. Furthermore, the budget

understates the total cost because it excludes indirect staff costs such as employee

18 http://www.doe.mass.edu/research/reports/2016/02METCO.pdf , p. 11.

7

benefits (health insurance and pension expenses). Figure 2’s left hand column shows the

budget as presented, while the right hand columns includes an adjustment to add

employee benefits , while redirecting an estimated portion of the support staff salaries 19

to bus monitors because these staff serve a dual role. Most importantly, even in the

unadjusted budget, no money is available to offset general education. With these

adjustments, it is clear that the grant funds even fail to cover the direct costs of METCO

programming, administration, and transportation. 20

19 The 26.9% employee benefit rate is used from the Town of Lexington, Financial Summit I (https://goo.gl/FoIaMh), 10/6/2016, p18. This report assumes non-tutoring METCO staff members are eligible for benefits. 20 The related paper on the financial history of METCO demonstrates that in past years METCO grants also covered the cost of teachers and materials. With the state setting the per pupil grant rate at such a low level, receiving districts are now essentially underwriting the full cost of classroom education and special education for METCO students.

8

Figure 2

A second, more detailed view of Lexington’s METCO expenses can be obtained by

reviewing the Lexington grant application to Massachusetts for Fund Code 317. Using 21

the revised total of $1.56 million the grant application further details the planned use of

funds. Again, in the grant application, no funds are proposed for common teachers and

materials.

21 http://www.doe.mass.edu/grants/2017/317/. Revised grant application for $1.56 million is used.

9

Figure 3. Lexington METCO Grant Application of Funds

10

As with the prior figure, two adjustments to this budget are proposed: estimating

employee benefits and allocating bus monitor expenses. The bus monitor allocation

makes clear that the Massachusetts METCO transportation grant amount of $525,094

falls about $150,000 short of covering transportation expenses in the Lexington METCO

program. Including employee benefits shows that Lexington spends more money on 22

METCO direct services than the full grant, the effect of excluding employee benefits from

school budgets. From these two illustrations, one can conclude that Lexington subsidizes

both transportation and direct services to students, and no funds are available to offset

indirect general education costs. Moreover, the METCO budget includes a number of 23

programs for which eligibility is based on METCO status, and not available to local

residents with equivalent educational need.

Full Operating Expense Impact of METCO Program

To estimate the net cost of the METCO program to Lexington, one casts the widest net to

include general state aid for students, while quantifying the cost of educating Boston

students. Chapter 70 aid is the principal vehicle for state revenue sharing with school

districts, providing Lexington an estimated $11.2 million in FY17. While Lexington

averages $1,600 per student, this report estimates that when METCO students are

treated as “last students in”, then an average $2,903 per student ($737K) is provided. 24

To quantify educational costs, this report uses an average cost approach. This average

cost is computed by subtracting the METCO program overhead from the school budget,

and then applying a base per pupil expenditure to METCO students. However, this

approach understates the operating cost of the METCO program as it does not quantify

METCO student over-representation in special education (SPED). 25

22 Lexington contracts with Metco Inc. to provide bus services, while Lexington pays for bus monitors who also serve as staff in the Lexington education program. 23 Lexington has funded a “late bus” in some years for METCO students to be able to participate in activities, and this may be an example of an unfunded cost believed necessary for METCO students to have some of the same extracurricular opportunities enjoyed by local students. 24 See appendix for detailed calculations. 25 In 2016-7, 19.4% of Lexington METCO students have an Individualized Education Plan (IEP), a subset of SPED programs (source: Lexington superintendent). In 2015, Lexington’s graduating cohort (inclusive of METCO students) had 11.0% IEPs (https://goo.gl/egTGf8). From these figures, one would estimate that Lexington residents have a 9-10% IEP rate. Other studies indicate that METCO students are identified at higher rates for SPED than suburban counterparts. See Pioneer Institute white paper Metco Merits More , 2011.

11

While some still advocate treating METCO students as “nearly free” because they fill

“empty seats”, such an argument fails for several reasons:

1. Lexington schools are operating close to school committee targets and

contractual maximums and therefore additional students trigger teacher aides or

increased classes. 26

2. METCO students are admitted to specific schools many months before final

enrollment numbers are known for a grade level, and influx rates are high enough

that one cannot predict where an “empty seat” would exist. 27

3. METCO students are not shifted across schools during or between years to “empty

seats” which become available.

4. Lexington’s school district is characterized by very high rates of in-migration

between kindergarten and grade 12, so an available seat in one year is unlikely to

be an available seat in the next year. In other words, if the third grade appeared 28

to have room for three students, those same three students might push a fourth

grade to have an extra classroom section. 29

In a single year, if one had sufficient data one could attempt to use a marginal cost

approach by identifying for each student whether they triggered an aid or teacher.

However, in the medium term, an average cost is more suitable because it accounts for

the capacity variations which occur from one year to the next. Using enrollment as a

cost driver, one should apply the economic principle that in the long run all costs are

variable , because an organization can determine how to restructure for efficiency at all 30

levels of scale. However, in the context of school administration and buildings historically

operating for a school population of 6000 students, there may also be significant

26 When Lincoln examined it’s METCO program in 2002, it’s class sizes averaged 16-20 across K-8 schools, far below Lexington’s average class size. Lincoln could add METCO students while offering class sizes well below those in Lexington today. Lexington contractual limits on class size are K: 20, Grade 1: 24, Grades 2-5: 26. https://goo.gl/TAzBOu, p11. 27 http://www.brooklinema.gov/DocumentCenter/Home/View/6346 , Brookline report indicates that METCO acceptance occurs in April of each year. If assignment to neighborhood schools is not guaranteed, one could also ask whether METCO students would be assigned to a specific school before or after residents moving into Lexington during the summer. 28 Lexington Ad Hoc Enrollment Working Group Progress Report, https://goo.gl/GKcw1D, slides 15-16. 29 Per contract, if the classroom maximum is exceeded an aide must be added, and if exceeded by 20%, the classroom must be split. https://goo.gl/TAzBOu, p11. 30 https://goo.gl/xUqSXh. Per pupil expenditure comparisons across school district suggest few scale economies of education.

12

diseconomies of scale causing any additional students (including METCO) to exceed the

average cost.

Figure 4 calculates the net METCO operating cost for the Lexington Public Schools.

● Per Pupil Expenditures (PPE) include benefits and maintenance (from state

reports)

● METCO enrollment is from the METCO grant application 31

● Chapter 70 Aid is estimated on a marginal basis (see Appendix for calculation).

● However, METCO expenses associated with elevated special education rates (and

the corresponding space footprint) are not included due to complexity and

inadequate data. 32

31 Grant is based on prior year enrollment, which is 254. As of this writing, the author has been told current year METCO enrollment may be 237. In this case, next year’s grant will be smaller. For simplicity the cost is shown as if the grant enrollment matches the next grant year, and these figures are never exactly the same. 32 One would next break out METCO students into special education (SPED) and non-SPED. Some METCO program funds might offset SPED use, while space usage would have to be detailed. METCO reports that their student population is identified at a 25% rate for SPED when placed in suburban programs, while one source reports that Newton METCO students have a SPED rate of 35%. (Phone conversation with Joshua Norman, 8/8/2016.) Performing this calculation exactly is difficult, as METCO students might continue to have higher SPED rates than the general Lexington population, but Boston students are not supposed to be eligible for Lexington’s expensive out-of-district SPED placements, placements which have partially offsetting state funds. A further nuance is that Lexington parents might push for out-of-district placement, while METCO parents might oppose out-of-district placement and remain at significant expense in district to avoid being returned to Boston. As an example, www.doemass.org/bsea/decisions/11-4687.doc details Lincoln bringing suit against Boston and a family to ensure an out of district placement over the family’s objection. This leads to bias in SPED representation as well as Lexington legal costs for METCO, another topic for which this report has no data. Report of the Lincoln Schools K-8 Task Force (2002) provides a more detailed calculation, identifying resident and METCO participation by SPED service level (out-of-district, high, moderate, and low). In that study, 26 out of 89 Lincoln METCO students received SPED services (29%), and METCO students accounted for 26% of all moderate service SPED and 25% of all high service SPED students in Lincoln.

13

Figure 4. Net METCO Operating Cost for Lexington Public Schools33

In total, Figure 4 shows that Lexington’s annual operating subsidy of METCO is about $4.2

million. Because Lexington schools are operating at or above design capacity both as

physical facilities and educational institutions, at the building level and in the classroom,

one should expect all students to have similar costs. However, this figure

underestimates the total cost of METCO because increased SPED participation is not

quantified, and because diseconomies of scale are likely occurring in Lexington schools.

These estimates are hard dollars, and do not include the educational impact of adding

more students to classrooms. If additional students simply occupied “empty seats” with

no educational impact, parents and educators would not labor tirelessly to keep

33 The total educational expenses reported as per pupil expenditures include both general fund expenses and grants and revolving funds (such as METCO or special education circuit breakers). Therefore, the $134 million reported here appears larger than similar figures reported on town general fund reports, such as the Finance Summit I document page 16 (https://goo.gl/FoIaMh).

14

classroom sizes ot a minimum. With many classrooms at or above target levels,

Lexington is surely experiencing an educational impact from crowding as well. 34

Lexington Capital Expenditures and METCO

The Lexington School Committee and other town committees are working to identify the

best options between acquiring new land, replacing schools, and renovating existing

schools. From a capital planning view, not only are there no “empty seats” in Lexington’s

schools, but many facilities and spaces are below current standards, and short term fixes

are patched into existing facilities due to critical needs and timing.

Because capital projects vary in scope and purpose, it is difficult to estimate precisely the

capital costs for Lexington’s METCO program. This report estimates the implied capital

costs from recent projects which do not involve land acquisition. This median annual

cost ($5,347 per pupil) is used for gross estimation, although reasonable figures could

vary from $2,200-$9,000 per pupil depending on circumstances. Based on this median

figure, the implied capital subsidy is about $1.4 million every year for Lexington’s METCO

students. 35

Figure 5. Lexington Capital Cost for METCO Program (excluding land)

34 FY17 Snapshot of Enrollments by Grade/School shows that all fifth grade classrooms in three of Lexington’s six elementary schools have 25-26 students, above the maximum target of 24 and at the contractual limit of 26 without an aid. As new students arrive during the year, these classrooms may be forced to add aids, or split. In one year, 34 new students joined this grade cohort, and one classroom was added. Many classrooms which were not crowded in one year became crowded in the next. 35 Details for projects assumptions are found in the Appendix.

15

Here it is worthwhile to note how critical assumptions factor into these calculations, and

reasons why the $1.4 million/year cost would be a conservative estimate:

● METCO students use special education (SPED) at higher rates, and SPED programs

require additional space.

● Land acquisition costs are excluded.

● Hastings costs are divided by all students, thereby driving a smaller cost per

student, although effective capacity has increased by only 222.

● A modest 2.5% interest rate is assumed for debt. 36

● Soft factors, such as space congestion, are not quantified.

Due to the continued enrollment surge, Lexington is evaluating land acquisition for

additional schools. Figure 6 estimates the cost of adding land acquisition costs to the

METCO capital costs projected in Figure 5. In this example, land costs are estimated at $1

million per acre , implying an additional annual charge of $1,051 per pupil, bringing the 37

total METCO capital impact to $1.6 million per year . This $1.6 million capital with land

figure is not used in the report summary, but included here to demonstrate additional

financial impact which Lexington’s METCO population has in the current school crowding

environment.

Figure 6. Lexington Capital Costs for METCO Program Including Land Acquisition

36 2.5% is national average rate available for AAA municipal bonds in the market as of 10/29/16, https://goo.gl/aUjEFS. The total cost of borrowing may be somewhat higher. 37 The Armenian Sisters Academy on Pelham Road was estimated at $8 million for 8.4 acres, but the current value of this land may be higher. https://goo.gl/RWba9c. Another benchmark is the recent town purchase of the Liberty Mutual building and land for $4.3 million, which includes 2.77 acres. Lexington Minuteman , “Town purchases building”, 10/27/2016, pA1.

16

Annual cost figures are useful to annualize METCO's cost to Lexington, but they

understate the equivalent capital expenditures. Building school capacity for 254 students

implies building projects valued in the tens of millions of dollars, and such projects are

material to Lexington’s continued AAA rating and overall debt service Today, Lexington 38

taxpayers are facing annual projects to add capacity (such as modulars at Fiske, Bowman,

and Bridge), projects which might be avoided or postponed if Boston students were not

enrolled.

Benchmark Studies from Other Communities

Researchers in peer communities have also substantiated the cost of METCO to receiving

districts. Despite variation in local conditions, recent estimates have been quite similar.

Figure 7 compares the operating loss associated with METCO across communities.

Figure 7: Evidence from METCO Community Studies39

Options to Close the Financial Gap

A just resolution of the financial gap would be for Boston to pay the operating and

capital costs for students sent to Lexington. However, in 1966 Boston did not sign on to

METCO due to its support for desegregation, but instead to discharge its educational

responsibility at the cost of other payers. The Boston School Committee voted 3-2 in

38 Total Lexington debt exclusions from 1990-2016 equal $181,377,248 (https://goo.gl/szoT1s). 39 Brookline: https://goo.gl/3IFqIH, Newton: Current estimate provided by Joshua Norman. Prior year’s estimates are online at https://goo.gl/j4uGAv (see out-of-district model), Lexington: this paper. Capital costs are also described in some studies, although presentation varies significantly.

17

support of the METCO plan in 1966 with the added condition that Boston never contribute

for the financial cost of these Boston residents . 40

Assuming Boston and Massachusetts will not close the financial gap, what are Lexington’s

options to close the $5.6 million per year gap and relieve overcrowding in Lexington’s

schools?

Option A: Phase Out or Eliminate

A four year phase out can be accomplished allowing most students to graduate their

current school.

Pros:

● Near term financial benefit, reducing pressure for operating override ● Reduces pressure on school capacity and capital planning by 250 students (with

associated SPED space). ● Informs capital planning, likely avoiding near term capital expenditures ● Phase out limits disruption to current METCO students

Cons:

● None

Precedents:

● Milton, Rockland, Framingham have withdrawn from METCO.

● Winchester, Andover, Reading, North Reading, Acton, Westford have no METCO

program.

Option B: Downsize

Reduce METCO student population from 250 to 100 students.

Pros:

● Limited financial benefit related to reduced use of teachers and materials ● Reduces school capacity needs by 150 students (with associated SPED space), with

limited potential for reduction of capital expenditures

40 Don’t Blame Us , Lily Geismer, p311, footnote 56.

18

● Signals to Massachusetts legislators the need for program funding

Cons:

● Continues violation of the Racial Imbalance Act 41

Precedents:

● Lincoln has downsized METCO considerably from peak ratios.

Option C: Reallocate METCO Grant Funds

Use state grant funds to offset teaching expenses, treating METCO students the same as

all other Lexington students by eliminating all administrative positions and specialized

programs.

Pros:

● At most $897,800 operating expenses available for alternative use. ● Reduces administrative overhead. ● Signals to Massachusetts legislature that desegregation requires significant

funding

Cons:.

● Large financial gap remains. ● No improvement on school crowding or capital projects. ● Some argue this option is impractical, as funds would still need to be used for

underachieving students, or the majority of METCO students could suffer academically. 42

● METCO has threatened to withdraw from districts which reduce spending. 43

● Continues violation of the Racial Imbalance Act.

41 Massachusetts Racial Imbalance Act violations are discussed in a related paper, How Lexington’s Voluntary Desegregation (METCO) Program Violates the Massachusetts Racial Imbalance Act (RIA) , by Mark Andersen, October 2016. 42 Past METCO student performance in Lexington has been poor. Jan 5, 1988: Over ⅓ of METCO middle school students earned D-F grades; Jan 6, 1988: over 63% of METCO high school students D-F grades in major course; Dec 21, 1987: 85% earn C- to F in high school. METCO Archives, Box 129, Folder 53, Northeastern University. Six METCO specific programs should be reviewed to determine if eligibility should be broadened to include Lexington residents with similar educational needs: Jumpstart Summer School, MELP, METCO middle school extended day, Mathpath summer school, METCO learning center, and METCO homework club. See Narrowing Achievement Gaps in Lexington Public Schools , Ferguson et. al., June 2015 43 “Metco parents vote Concord-Carlisle High Withdrawal”, Boston Globe , July 13, 1978, p3.

19

Precedents:

● Framingham, Concord-Carlisle, and Newton attempted reductions although only Framingham followed through completely. Newton’s threatened reduction resulted in Governor Dukakis providing additional funds. Lincoln’s review of 44

METCO-specific resources led to extending full day kindergarten from METCO-only to its own residents. 45

Alternative Uses for $5.6 Million Per Year

One use of $5.6 million per year would be to reduce taxes. Property tax revenue for

Lexington in 2016 are $154.8 million , so this would be equal to a 3.6% reduction in taxes 46

for all households. Distributed across 10,195 households, this is equivalent to an average

$549 per household per year .

An alternative use for $5.6 million would be to invest in educational or town programs.

On the educational side, the Lexington School Committee has expressed interest in

elementary foreign language, but has been unable to fund the $2.3 million per year

required. A second possibility would be to expand the mathematics curriculum,

responding to the estimated 1,000 students (1 in 7) who are engaged in privately funded

mathematics programs outside the school day.

On the town side, $5.6 million per year over 30 years is equal to about a $118,000,000

capital investment (at 2.5% interest rate). In other words, the town might use the savings

to fund a series of planned projects: a new police station, a new fire station, and a new

visitor center. It is also possible to apply these funds to close the $25 million gap in 47

Lexington’s unfunded pension liabilities. Or with MSBA support, these funds could pay 48

for half of a new high school, updated to current expectations for science and

engineering labs.

44 Geismer, Don’t Blame Us , cites Patriot Ledger August 28, 1976, “Dukakis shifting $150,000 to keep Newton in METCO”, p221. 45 A Qualitative Examination of Desegregated Education Through the METCO Program , Kahris Dianne White-McLaughlin, 1996, Lesley College, p285-7. 46 FY2018 Summit I Indicators, October 6, 2016. 47 Fire station costs: https://goo.gl/IJRHRD , Police stations: https://goo.gl/ZPZLPQ. The Lexington fire station is estimated to cost $19 million, Lexington Minuteman , “Town purchases building”, 10/27/16, pA8. 48 Financial Summit I , p20.

20

The bottom line is that $5.6 million in a single year might seem sizable, but a

commitment to spend $5.6 million per year on an ongoing basis is an enormous financial

liability. At a 4% discount rate, this liability has a net present value of $140 million.

Benchmarking Lexington’s METCO Program Size

In determining how to proceed, some residents have asked about the comparative size

of Lexington’s METCO program. One way to benchmark Lexington’s METCO program is

to compare it to other Boston suburbs (figure 8). Lexington’s METCO program is the third

largest in Massachusetts with an enrollment of 254, only surpassed by Newton and

Brookline. Only 29 Boston suburbs participate directly in METCO out of 100 cities and 49

towns in the Boston Metropolitan Area Planning Council. Lexington accepts 8% of all

Boston area METCO students , yet Lexington’s 31,394 residents account for less than 1% of

the Boston Metropolitan Statistical Area population. Most Boston suburbs receive zero 50

students.

49 Lexington has fewer households supporting each METCO student than either Newton or Brookline, as 2010 Census figures indicate Brookline has 2.2 times as many households, and Newton has 2.7 times as many households. Higher household to METCO student ratios in Brookline and Newton suggest better ability to absorb the cost of education. 50 The Boston Metropolitan Statistical Area’s population of 4.7 million people, of which 645 thousand live in Boston. https://en.wikipedia.org/wiki/Greater_Boston.

21

Figure 8. METCO Program Size (Participating Districts)51

Contributing a Fair Share to Public Education

Some argue that Lexington should fund the METCO subsidy to be generous towards

Boston minority students. Financial analysis demonstrates that Lexington schools attract

a disproportionate number of students already, and therefore taxpayers already educate

far more than their “fair share” of public school students. One way to think about

51 In this and subsequent charts and the accompanying descriptions and rankings, regional districts are combined into a single unit (Lincoln-Sudbury, Concord-Carlisle, and Dover-Sherborn). Thus there are three “programs” listed, and ratios computed for these three, when in fact there may be as many as three programs per joint town. Carlisle does not participate in METCO directly in 2016. This chart could be extended to enumerate more than fifty towns with zero METCO participation. Figures are for October 2015 enrollment per the METCO grant.

22

fairness is how many adults are in a community per public school attending child. While

across Massachusetts there are 5.6 adults per public school K-12 student, in Lexington

the ratio is 3.4, among the lowest in the state (figure 9 shows Lexington as the seventh

among Boston area districts). If one only considered Lexington income levels, one would

have an incomplete story as town residents pay far above average taxation because

excellent schools create an educational burden (ratio) greater than Boston and nearly all

other communities. And while Lexington’s educational burden is greater than Boston’s,

and it’s tax rates considerably higher, it subsidizes Boston public education by funding

regular and special education for Boston students.

23

Figure 9: Adults Supporting Each K-12 Public School Student52

52 Ratio of 2010 population 18 years and older to 2015 public school K-12 children. Boston public school district is used for ratio (excludes charter schools).

24

A second metric is households per public school child. Again, Lexington is among a

handful of towns with fewer than 1.7 households per student (see figure 10 for Boston area

communities in which Lexington is sixth), while the state average is 2.75 households per

student. The beneficiary of the Lexington METCO subsidy, Boston, has five households

per public school student, among the most favorable in the region.

25

Figure 10: Households Supporting Each K-12 Public School Student53

53 Ratio of 2010 Census Households to 2015 public school K-12 children

26

These figures highlight that Lexington’s large and growing student population constitute

a disproportionate burden when compared with other Boston suburbs, with Boston itself

(gray line), or across the Commonwealth. Adding Lexington’s large METCO population to

an already high taxpayer burden cannot be justified on the basis of moral obligation for

generosity. Interestingly, the two larger METCO participants lie closer to average

household ratios ratios (Newton: 2.5, Brookline: 3.4). If Lexington did not have so many

public school aged children (26% of the population is under 18), it would be better

positioned to subsidize a substantial METCO program. Because Lexington has higher

public school attendance rates than Newton or Brookline, it does not have comparable

funds to educate Boston students. Moreover, Boston’s superior ratios and considerably 54

lower residential tax rates suggest that in 2016 Boston is better positioned than

Lexington to fund education of Boston’s METCO population.

Conclusion

METCO Inc. describes the program as costing receiving districts little because students

occupy so-called “empty seats”. The Brookline task force on Population and Special 55

Education effectively debunked the myth that METCO students occupy empty seats,

pointing out that these students are approved in spring, long before final enrollment

numbers are known, and once enrolled remain until graduation. 56

Lexington’s METCO program enrolls about 254 students, equal to 8% of the Boston

METCO population, making it the third largest METCO program in the state. Boston

contributes nothing. Massachusetts provides only a small grant to cover transportation

and most, but not all, of the cost of METCO direct services. Using this funding, Lexington

has offered at least six different educational programs to METCO students which are not

available to Lexington students with comparable educational needs. 57

54 Percentage of school aged children attending public school: Lexington: 90% Brookline: 87%, Newton: 81% public school. https://goo.gl/9wn5QK (2015-16). In Lexington, a 1% shift equals 75 students in Lexington, so a 3% shift to Brookline levels would nearly equal the entire METCO population. Lexington had had higher public school attendance rates in the past, for example 93% in 1995. 55 August 2016 version of METCO website. Also http://brookline.wickedlocal.com/article/20140611/news/140619261. 56 http://www.brooklinema.gov/DocumentCenter/Home/View/6346 p. iii, p19. 57 Narrowing Achievement Gaps in Lexington Public Schools , Ferguson et. al., June 2015. Earlier footnote has list of programs.

27

The net cost to Lexington taxpayers is estimated to be $5.6 million per year. Of this, $4.2

million per year are operating costs to provide teachers, counselors, administrators,

laboratories, and materials for METCO students. An additional $1.4 million per year is

the estimated capital expense. These costs are readily visible in the numerous projects

required to install modulars across the Lexington school district. These estimated costs

are conservative, as they ignore possible land acquisition for new schools, crowding

effects in classrooms and on playgrounds, diseconomies of scale at Lexington’s present

size, and disproportionate METCO special education needs.

Options for the Lexington School Committee to close this financial gap include phasing

out, downsizing, or eliminating expenses specific to METCO students. However, only

phasing out or substantially reducing Lexington METCO will contribute materially to

school crowding which has increased in recent years. The paper also compares the

public education burden on Lexington taxpayers, and finds that Lexington’s excellent

schools attract so many students that taxpayers bear an educational burden far larger

than other communities, and specifically Boston. Boston residents enjoy a lower tax rate,

while the Boston School Committee contributes zero funds to the regular and special

education, or even transportation, of its students educated in Lexington. In this time of

fiscal crisis driven by surging enrollments, the Lexington School Committee and

taxpayers will need to discuss the best options to secure funding or close the gap.

28

Appendices

Appendix I: Glossary and Abbreviations

METCO Inc. The Roxbury based organization which has coordinated services and

referred students for the METCO program since 1966. METCO stands for Metropolitan

Council for Educational Opportunity. The Massachusetts legislature specifies that METCO

Inc. administer the student referral process for the METCO program. 58

Racial Imbalance Act (RIA). The 1966 Act passed by the Massachusetts legislature,

amended in 1966 and again in 1974, which is the legal basis for mandatory and voluntary

desegregation in Massachusetts. This Act defines the terms Racial Balance, Racial

Imbalance, and Racial Isolation. Careful reading of this act suggests that Lexington might

not be an eligible receiver of METCO students due to Lexington’s own racial imbalance.

DoE. Massachusetts Department of Education administers the METCO grant funds.

MGL. Massachusetts General Laws

SPED. Special Education. Additional services offered to students who have been

identified to have additional needs.

ELL. English Language Learner. Students which are eligible for special programming

because they have not yet adjusted to teaching in English.

MSBA. Massachusetts School Building Authority. Reviews and contributes to select

school capital projects.

58 METCO Inc. has administered the program without competitive bid for fifty years.

29

Appendix II: METCO Size Relative To School Population

The main body of this report benchmarked schools based on total METCO student

population. An alternative benchmark for program size is the ratio of resident pupils to

METCO pupils. According to this measure, Lexington has the seventh largest ratio (1:26)

in the Boston metropolitan area, again out of 100 Boston area communities. However,

this benchmark is less useful, because Lexington’s METCO program size appears diluted

with each additional resident student. Combined with the principal benchmark (program

size), Lexington has a much larger program than an average participating district, and far

larger than the average Boston regional suburb (which does not even participate in

METCO).

Figure 11. METCO Student to Public K:12 Ratio (for Boston suburbs with METCO)

30

Figure 11 shows the ratio between public K-12 pupils and METCO enrollment, with larger bars representing

proportionally larger METCO programs. Lexington’s 1:26 ratio means that is one METCO student for every 26

resident Lexington students.

Appendix III: Lexington School Capital Projects

This appendix substantiates the assumptions used in estimating the cost and capacity

effect for Lexington’s capital school projects. As project costs vary by size, scope and

purpose, the purpose of this report is to standardize to provide a benchmark for capital

costs.

Today’s Lexington’s school projects are seldom “first-class”, but instead reflect frugality

and an attempt to apply space mining or modular additions to accommodate a rapidly

growing school population. The only exception are Massachusetts School Building

Authority (MSBA) subsidized projects (such as Hastings) where state standards dictate

many parameters of size. These capital expansion do not include land, might not pay for

the original building envelope, and do not reflect the congestion effects of overburdened

classrooms, hallways, and playgrounds. Disregarding these factors, and applying

conservative methods for cost assumptions, this report likely underestimates the capital

impact of METCO students. 59

Elementary Projects: Hastings

Due to a large boom in Lexington enrollments, Hastings elementary school is now slated

to be torn down and reconstructed as a larger school, with a similar footprint to the

recently constructed Estabrook. The Statement of Interest for Hastings indicates it will cost

about $60 million to build, with the hope that the state might come forward with as much

as $20 million leaving Lexington taxpayers with only a $40 million bill. The article 2 60

taxpayer impact handout cites the final Lexington figure as $42 million. 61

59 Maintenance cost is part of per pupil expenditures, and is not included in capital estimation here. 60 https://goo.gl/dH4iDM, accessed 8/4/2016. 61 Copy of Article 2 provided to special town meeting and sent to this author.

31

From a pure capacity standpoint, the Hastings project will increase enrollment from 423

students (2014-2015) to a MSBA design certification of 645 students, thus adding 222 62

students. Using 222 students as a denominator would point to a large cost for

incremental capacity. Because Hastings replacement has long been part of the master

plan, this report conservatively divides the build cost (minus MSBA funds) over 645

students. Land is excluded (Hastings school remains at 14 acres), and congestion effects

are ignored.

Elementary Projects: Modulars at Bridge, Bowman and Fiske

Crowding conditions exist at all elementary schools in Lexington except for Estabrook.

This crowding is created by a combination of increased enrollment and high space

demands for specialized space (such as growth in English-Language Learner (ELL)

programs in Lexington). In 2016, Lexington appropriated $2.92 million for modulars at 63

these three schools, which is currently estimated to purchase five classrooms. However,

this space provides relief for extreme crowding and does not increase student capacity,

and therefore these examples do not lend themselves to estimating the cost for capacity.

It is noteworthy that these additional costs are tied to specialized programs, some of

which METCO may use in disproportionate numbers.

Middle School Projects

In 2016, the Lexington School Committee, Town Meeting, and voters approved space

mining and expansion for the two Lexington middle schools: Diamond and Clarke. The

final appropriation for this expansion is $66.6 million ($44.94 million for Diamond and

$21.68 million for Clarke). Subtracting $10 million for deferred replacement of the HVAC

system at Diamond, one estimates the net $56.6 million as intended to increase in

student capacity. 64

62 https://dl.dropboxusercontent.com/u/321011924/Lexington%20Multiple%20Projects/SMMA%20Masterplan%20LexingtonReport_10FEB2015.pdf, p278. 63 http://lexingtontmma.org/uploads/Main/WarrantInfoFall2015STM1V1.pdf 64 $10 million HVAC replacement pointed out by Pat Goddard, conversation 8/5/2016. Pat also provided insights helpful to these initial estimates, for which I take responsibility.

32

How much additional capacity is made available for this sum? In 2015-2016, Clarke had

864 students and Diamond 778, for a total 1,642 students. When the project is complete,

Diamond will have increased capacity from 9 teams to 11 teams and Clarke will have a

capacity of 10.5 teams. With a total of 21.5 teams at an educational target of 86 students

per team, this equals 1,849 students, an increase of 207 students. However, this planned

capacity might be ultimately exceeded, so this report assumes more conservatively a

higher figure of 90 students per team as a rated capacity, thereby pointing to a potential

capacity of 1,935 students across two schools, an increase of 297 students from the

2015-2016 level. 65

Lexington High School

The Whole Building Design Guide (WBDG) website describes the 2014 high school

prefabrication project. For $4.9 million, 10 new classrooms and some special education 66

spaces were added. As prefab buildings, the Lexington building department assumes 20

years useful life. Prior to construction, LHS had 2002 students (2013-2014). With the 67

addition of these modulars, LHS then supported 2093 students in 2014 and 2154

students in 2015. Apparently 2154 was the target capacity with these classrooms added.

Therefore, one might estimate that capacity for 152 students was added, which seems to

correspond roughly to the 10 classrooms plus special education spaces. Once again, this

cost does not take into account land nor congestion, such as the loss of the freshman

quad at LHS in which the prefab was placed.

Minuteman Vocational-Technical High School

In September 2016, voters approved a new building for Minuteman Vocational-Technical

High School. This example is useful both because it represents a significant high school

project, and the state approved an explicit capital charge for out-of-district students was

calculated for Minuteman at $5953/pupil. It is precisely this out-of-district charge that

65 As in all these analyses, this is not a total cost, as the land cost and land congestion effects are disregarded. Master planning and some design costs are also not incorporated. 66 https://www.wbdg.org/references/cs_lexingtonhs.php, referenced 8/4/2016. 67 Director of Public Facilities, Pat Goddard stated that the expected lifetime is 20 years, while the Whole Building Design Guide site indicates 10-15 years for prefab construction. This report uses 20 years to be conservative and consistent with Lexington’s history of using school buildings for longer than expected life. Conversation 8/5/2016

33

Lexington should also calculate for its own buildings, so it can make optimal decisions for

resource allocation in cases such as METCO. Arguably Minuteman may have some

specialized facilities beyond those required for METCO students, although due to higher

SPED rates among METCO students it is possible that the capital cost per METCO student

is higher than per vocational-technical high school student.

34

Appendix IV: Chapter 70 Foundation Aid Impact

Since McDuffy v. Secretary of the Executive Office of Education (1993) and the 1993 Education

Reform Act, the state has used a revised foundation formula for education aid to cities

and towns. Unfortunately, calculating Chapter 70 aid is incredibly complicated, and best

understood by simulating enrollment with the state worksheet. Simply dividing $11.17 68

million aid by 6,978 students shows an average aid per student of $1,600, but this does

not reflect two special classes of students, nor the fact that the first student receives little

aid while the last one receives the most.

Two classes of students merit increased aid: students enrolled in English Language

Learner (ELL) programs, and starting in FY17, economically disadvantaged students. 69

The state formula provides SPED reimbursement as well, but does not vary by actual

SPED enrollment, but instead uses the (absurdly) low fixed figure of 3.75% of pupils in

SPED. Therefore, populations with above average SPED participation rates (such as 70

METCO students) do not receive additional Chapter 70 aid.

Without exact figures for Lexington’s METCO population, this report uses METCO-wide

figures from METCO Merits More to estimate economically disadvantaged and ELL 71

students. Using that paper’s 50% low-income and 2.6% ELL figure, the author estimates

that 126 METCO students may be low-income and 7 enrolled in ELL. Using the state

calculation worksheet, the author removed 254 pupils from the system, testing three

cases: one with all economically disadvantaged, a second with none disadvantaged, and a

third with an estimated 126 disadvantaged (METCO scenario). The results are shown in

figure 12. 72

68 All calculations done in this section are based on the state formula spreadsheet which can be found at: https://goo.gl/u5CXQ1. The total # of students does not seem to match exactly the sum of subcategories. A whitepaper discusses the calculation and confirms that METCO students are included: https://goo.gl/6cQfXc 69 Disadvantaged students are identified by “participation in one or more of the following state-administered programs: Supplemental Nutrition Assistance Program (SNAP); Transitional Assistance for Families with Dependent Children (TAFDC); Department of Children and Families' (DCF) foster care program; or MassHealth (Medicaid) up to 133% of the federal poverty level (FPL).” Ibid ., p5. 70 “This method is in place to prevent over identification of special education students for the purpose of fiscal gain”, Ibid., p5. 71 METCO Merits More , p.9,11. 72 The following estimates were used for the state worksheet: 19 full-day K, 96 elementary, 57 middle school, 75 high school, and 7 ELL students. There seem to be small discrepancies in total student

35

Figure 12: Chapter 70 Aid with Hypothetical Adjustments

Given the complexity of the spreadsheet, these calculations should be taken as a best

approximation for Chapter 70 aid impact. Key observations can be made:

● Estimate that if 254 METCO students were removed, Lexington would receive

$737,487 less Chapter 70 aid.

● Marginal (last) students are awarded more aid than average students,

approximately $3768 per economically disadvantaged pupil, and $2052 for

unclassified pupil. 73

● The other columns provide sensitivity analysis, suggesting that actual values could

vary by +/- $100,000, which is small in the context of METCO costs.

Based on this model, a total amount of $737,487 is estimated to be related to the

education of METCO pupils. This figure is used in the main operating cost calculation.

The Pioneer Institute white paper Metco Merits More states about Chapter 70 aid:

Even with the addition of Chapter 70 aid, though, METCO-related dollars (typically

$4,900 per student for a METCO grant plus transportation allocation) coming into

receiving districts does not begin to match the per-pupil expenditures either in the

suburbs or urban areas. 74

counts in the state spreadsheet, for example with 24 fewer students in the 6.978 total than the sum of pre-K through ELL (7,001). The author has done his best to provide an estimate despite these issues. At the fixed 3.75% SPED rate, removing 254 METCO students causes 9 fewer SPED students in the formula. 73 Note: this calculation treats the METCO students as the last pupils to be added or subtracted from the district, thereby providing the maximum figures per student. 74 Metco Merits More , p12.

36